The optimal pattern of cancer care for patients failing standard therapy remains unsettled. Three paths are now available in the United States for terminal care: 1. Hospitalization, 2. Hospice or chronic care institutionalization, or 3. Home care. This study is designed to evaluate the choice of hospitalization vs. home care with optimal medical support to determine whether: 1. Home care is cost-effective. 2. Psychologic adjustment to the illness by both family and patient are modified by pattern of care. 3. Survival and quality of life is enhanced by home care. 4. Narcotic usage is modified by environment. Patients with evidence of progressive cancer in spite of optimal medical therapy will be studied prospectively by a physician, social worker, and dietician to determine whether or not home care is beneficial. Standard techniques such as the Karnofsky Index, profile of mood states, multiple affect adjective check list and caloric count will be applied to equivalent groups of patients who recieve medical support either in the hospital or at home.